prometheus.exporter.postgres
The prometheus.exporter.postgres
component embeds the postgres_exporter
for collecting metrics from a PostgreSQL database.
Multiple prometheus.exporter.postgres
components can be specified by giving them different labels.
Usage
prometheus.exporter.postgres "LABEL" {
data_source_names = DATA_SOURCE_NAMES_LIST
}
Arguments
The following arguments are supported:
Name | Type | Description | Default | Required |
---|---|---|---|---|
data_source_names | list(secret) | Specifies the PostgreSQL servers to connect to. | yes | |
disable_settings_metrics | bool | Disables collection of metrics from pg_settings . | false | no |
disable_default_metrics | bool | When true , only exposes metrics supplied from custom_queries_config_path . | false | no |
custom_queries_config_path | string | Path to YAML file containing custom queries to expose as metrics. | "" | no |
enabled_collectors | list(string) | List of collectors to enable. Refer to the information below for more detail. | [] | no |
Refer to the PostgreSQL documentation for more information about the format of the connection strings in data_source_names
.
Refer to the examples for the custom_queries_config_path
file in the postgres_exporter
repository.
Note
There are a number of environment variables that aren’t recommended for use, as they will affect allprometheus.exporter.postgres
components. Refer to thepostgres_exporter
repository for a full list of environment variables.
By default, the same set of metrics is enabled as in the upstream postgres_exporter
.
If custom_queries_config_path
is set, additional metrics defined in the given configuration file will be exposed.
If disable_default_metrics
is set to true
, only the metrics defined in the custom_queries_config_path
file will be exposed.
A subset of metrics collectors can be controlled by setting the enabled_collectors
argument.
The following collectors are available for selection:
database
database_wraparound
locks
long_running_transactions
postmaster
process_idle
replication
replication_slot
stat_activity_autovacuum
stat_bgwriter
stat_database
stat_statements
stat_user_tables
stat_wal_receiver
statio_user_indexes
statio_user_tables
wal
xlog_location
By default, the following collectors are enabled:
database
locks
replication
replication_slot
stat_bgwriter
stat_database
stat_user_tables
statio_user_tables
wal
Note
Due to a limitation of the upstream exporter, when multipledata_source_names
are used, the collectors that are controlled via theenabled_collectors
argument will only be applied to the first data source in the list.
Blocks
The following blocks are supported:
Hierarchy | Block | Description | Required |
---|---|---|---|
autodiscovery | autodiscovery | Database discovery settings. | no |
autodiscovery block
The autodiscovery
block configures discovery of databases, outside of any specified in data_source_names
.
The following arguments are supported:
Name | Type | Description | Default | Required |
---|---|---|---|---|
enabled | bool | Whether to automatically discover other databases. | false | no |
database_allowlist | list(string) | List of databases to filter for, meaning only these databases will be scraped. | no | |
database_denylist | list(string) | List of databases to filter out, meaning all other databases will be scraped. | no |
If enabled
is set to true
and no allowlist or denylist is specified, the exporter will scrape from all databases.
If autodiscovery
is disabled, neither database_allowlist
nor database_denylist
will have any effect.
Exported fields
The following fields are exported and can be referenced by other components.
Name | Type | Description |
---|---|---|
targets | list(map(string)) | The targets that can be used to collect exporter metrics. |
For example, the targets
can either be passed to a discovery.relabel
component to rewrite the targets’ label sets or to a prometheus.scrape
component that collects the exposed metrics.
The exported targets use the configured in-memory traffic address specified by the run command.
Component health
prometheus.exporter.postgres
is only reported as unhealthy if given an invalid configuration.
Debug information
prometheus.exporter.postgres
doesn’t expose any component-specific debug information.
Debug metrics
prometheus.exporter.postgres
doesn’t expose any component-specific debug metrics.
Examples
Collect metrics from a PostgreSQL server
This example uses a prometheus.exporter.postgres
component to collect metrics from a PostgreSQL server running locally with all default settings:
// Because no autodiscovery is defined, this will only scrape the 'database_name' database, as defined
// in the DSN below.
prometheus.exporter.postgres "example" {
data_source_names = ["postgresql://username:password@localhost:5432/database_name?sslmode=disable"]
}
prometheus.scrape "default" {
targets = prometheus.exporter.postgres.example.targets
forward_to = [prometheus.remote_write.demo.receiver]
}
prometheus.remote_write "demo" {
endpoint {
url = PROMETHEUS_REMOTE_WRITE_URL
basic_auth {
username = USERNAME
password = PASSWORD
}
}
}
Replace the following:
PROMETHEUS_REMOTE_WRITE_URL
: The URL of the Prometheus remote_write-compatible server to send metrics to.USERNAME
: The username to use for authentication to theremote_write
API.PASSWORD
: The password to use for authentication to theremote_write
API.
Collect custom metrics from an allowlisted set of databases
This example uses a prometheus.exporter.postgres
component to collect custom metrics from a set of
specific databases, replacing default metrics with custom metrics derived from queries in /etc/alloy/custom-postgres-metrics.yaml
:
prometheus.exporter.postgres "example" {
data_source_names = ["postgresql://username:password@localhost:5432/database_name?sslmode=disable"]
// This block configures autodiscovery to check for databases outside of the 'database_name' db
// specified in the DSN above. The database_allowlist field means that only the 'frontend_app' and 'backend_app'
// databases will be scraped.
autodiscovery {
enabled = true
database_allowlist = ["frontend_app", "backend_app"]
}
disable_default_metrics = true
custom_queries_config_path = "/etc/alloy/custom-postgres-metrics.yaml"
}
prometheus.scrape "default" {
targets = prometheus.exporter.postgres.example.targets
forward_to = [prometheus.remote_write.demo.receiver]
}
prometheus.remote_write "demo" {
endpoint {
url = PROMETHEUS_REMOTE_WRITE_URL
basic_auth {
username = USERNAME
password = PASSWORD
}
}
}
Replace the following:
PROMETHEUS_REMOTE_WRITE_URL
: The URL of the Prometheus remote_write-compatible server to send metrics to.USERNAME
: The username to use for authentication to theremote_write
API.PASSWORD
: The password to use for authentication to theremote_write
API.
Collect metrics from all databases except for a denylisted database
This example uses a prometheus.exporter.postgres
component to collect custom metrics from all databases except for the secrets
database.
prometheus.exporter.postgres "example" {
data_source_names = ["postgresql://username:password@localhost:5432/database_name?sslmode=disable"]
// The database_denylist field will filter out those databases from the list of databases to scrape,
// meaning that all databases *except* these will be scraped.
//
// In this example it will scrape all databases except for the one named 'secrets'.
autodiscovery {
enabled = true
database_denylist = ["secrets"]
}
}
prometheus.scrape "default" {
targets = prometheus.exporter.postgres.example.targets
forward_to = [prometheus.remote_write.demo.receiver]
}
prometheus.remote_write "demo" {
endpoint {
url = PROMETHEUS_REMOTE_WRITE_URL
basic_auth {
username = USERNAME
password = PASSWORD
}
}
}
Replace the following:
PROMETHEUS_REMOTE_WRITE_URL
: The URL of the Prometheus remote_write-compatible server to send metrics to.USERNAME
: The username to use for authentication to theremote_write
API.PASSWORD
: The password to use for authentication to theremote_write
API.
Compatible components
prometheus.exporter.postgres
has exports that can be consumed by the following components:
- Components that consume Targets
Note
Connecting some components may not be sensible or components may require further configuration to make the connection work correctly. Refer to the linked documentation for more details.